Gender analysis gives new perspectives in research

What are the consequences when drugs or airbags are only tested on men? Or when robots are programmed to maintain stereotypes?

Without the gender dimension, research risks disadvantaging large groups of people according to Czech researcher Marcela Linkova, Director of the Center for Gender and Science at the Czech Academy of Sciences.

“The integration of a gender dimension into research and innovation means better studies, better solutions. Research funding bodies must take their share of responsibility in relation to the community at large, and ensure that research generates answers that benefit a range of groups,” she says.

She sees big differences between the countries in Europe:
“In some countries, research councils have moved quickly forward. But in many others, unfortunately this issue is not on the agenda at all,” she says.

The need for the gender dimension in fields such as medicine and technology is particularly urgent, according to Londa Schiebinger, Professor of History of Science at Stanford University in the USA. She is a world leader in the field of gender in science and innovation. In the last 35 years, she has advised researchers as well as decision-makers about the importance of including a gender dimension in scientific research.

“It is still the case that medical research is done primarily on males - whether cells, animals, or humans. Doing the research wrong can cost lives and money,” she says.

For example she mentions that 8 of the 10 drugs withdrawn from the US market because of life-threatening health effects posed greater health risks for women than for men.

International collaboration highlights why gender matters

Londa Schiebinger leads an international collaboration entitled “Gendered Innovations,” which highlights concrete illustrations of how sex and gender analysis lead to discovery and innovation. Many examples of research are described at the website genderedinnovations.com

Over the past few decades, knowledge has increased, but Londa Schiebinger still sees big gaps, in particular in the area of technology. For example, crash tests for new cars often only use dummies made to mimic an average 45-year-old man. Women are generally shorter than men and the deficiencies in these safety tests could be one reason why women suffer whiplash injuries at double the frequency of men in traffic accidents.

“When we develop cars, we must make sure that everyone is taken into consideration. We need to ensure that research benefits all people,” says Londa Schiebinger.

As a positive example, she mentions that already in 2002 Volvo produced a virtual crash test dummy made to mimic a pregnant woman. Londa Schiebinger says it is crucial for car manufacturers and national safety agencies to study how different body types are impacted by traffic accidents.

She also stresses the need for integrating the gender dimension into the development of artificial intelligence (AI) and the new assistive technologies. There is a risk that artificial intelligence will pick up and amplify human social biases, she explains.

“In Google Search, for example, men are five times more likely than women to be offered ads for high-paying executive jobs. No one meant this to happen; it is unconscious bias. The bias already existed in the data the that algorithm was trained on,” says Londa Schiebinger.

The company Amazon encountered a similar problem when they wanted to develop a smart recruitment tool. The idea was that the AI should go through the applications received and select the best candidates. But because of the algorithm behind the, men were more likely to be selected for jobs in technology. The program learned based on who had been employed in the past and because this was mainly men, the AI assumed that men would be the more desirable candidates.

Researchers must consider the gender dimension

British researcher Elizabeth Pollitzer leads an organization called Portia, which is working for the inclusion of the gender dimension in STEM subjects (Science, Technology, Engineering and Mathematics). She thinks that research councils have an important role to play and wants more of them to routinely require researchers to show whether a gender dimension is relevant in their projects. In her view, it should be mandatory in all research applications.

“It’s a matter of ensuring that you get good quality research that is applicable across the whole of society. Ultimately, the research is financed by all of us through our taxes, and that makes it doubly relevant that it should apply to everyone,” she says.

Since 2011, as an outcome from an EU-funded project, Portia has arranged the international conference entitled Gender Summit. The conference is a platform for collaboration between researchers, financiers, decision-makers and other stakeholders. Elizabeth Pollitzer says that internationally there is great commitment to the inclusion of the gender dimension in research, but there is also some resistance.

“It’s a contentious issue because it’s about changing the culture in science. We are questioning the way in which research has been done previously, and what is required for research to be deemed excellent,” she says.
What she sees is that academic journals generally place comparatively higher demands on the integration of the gender dimension than the research councils. Many journals require that researchers account for whether a gender dimension is relevant or not – and why.

“I think that the research councils should do the same. It would be good if the journals and the research councils could come up with and establish a common standard. That would make it easier for both analysts and researchers to know what is expected,” she says.

Risk of setbacks in the EU

Elizabeth Pollitzer identifies the EU as an important role model in Europe. When the EU highlighted the importance of the gender dimension in its research programme Horizon 2020, many European research councils followed suit. But whether or not the EU will continue to lead the way in the future is uncertain, according to Pollitzer. Horizon 2020 ends next year and will be replaced by a new programme called Horizon Europe, which is currently being developed.

“As things stand now, the gender dimension is not going to be given the same weight at all as it had before. There is no advisory group on the gender area, and which research topics are to include the gender dimension is not at all as clearly described as previously. I am really worried about how it’s going to turn out. All work on the integration of the gender dimension in EU-funded research could disappear,” says Pollitzer.

She emphasises the importance of scientists including a gender dimension right from the start of a project.

“It affects how scientists pose their questions and what data is gathered. You can’t just do a gender analysis at the end,” she says.

By now, most researchers and funding bodies ought to know this, according to Marcela Linkova.

“From history, we know that if we don’t include the gender dimension, it can result in research that turns out to be harmful or even fatal,” she says.

But she also stresses that researchers should not overly focus on gender, as for example some medical researchers in neurology have done. Linkova is referring to studies where researchers have found differences between men’s and women’s brains and from these drawn far-reaching conclusions about differences in cognitive ability.

“There seems to be no end to this in spite of the fact that we have scientific evidence to prove that it’s not true,” she says.

The Gendered Innovations website gives another example of how things can turn out when researchers exaggerate the importance of sex differences. It describes how a medical knee replacement project produced a special type of knee replacement adapted for women, despite the fact that, in this case, it might be better to base the implant on other factors, such as the patient’s height.

“Sex and gender are two dimensions among many others. Sometimes they will be important, sometimes other factors will be more important. You need to include a variety of variables and perspectives in the design of research in order to get it right,” says Londa Schiebinger.

The gender dimension is missing in education

Londa Schiebinger thinks that universities have an important role to play in educating researchers and future researchers in what it means to include sex or gender analysis as potential variables in their research.

“I don’t know of any university that does a good job of integrating a gender dimension into core courses in IT. There need to be workshops and courses to make researchers and markers more aware of what it means to integrate sex or gender into the analysis,” she says.

Sometimes the issue of integrating the gender dimension into research gets mixed up with work to improve gender equality within academia. The Gendered Innovation website describes gender equality as a matter of “fixing the numbers” and “fixing the institutions”. It’s about ensuring that women in academia have the same opportunities to pursue a career as men. The integration of the gender dimension into research is about “fixing the knowledge”.

There is no definite connection between increasing the proportion of women in research and integrating the gender dimension into research,” says Marcela Linkova.
She describes them as separate – but related – issues.

“Both women and men can be gender-blind, but we know that there are advantages with groups that are not homogeneous. More women and greater ethnic diversity is likely to mean that we get to include questions and insights that would not otherwise have come up,” she says.

Text by Charlie Olofsson